According to a long-held doctrine, no significant numbers of neurons are made in the adult mammal to contribute to the function of the adult mammalian nervous system. However, recent data indicate that adult mammalian brains contain neural precursor cells capable of generating new neurons both in normal and in injured conditions. These new neurons have been quantified in live animals by injecting or feeding in drinking water a marker of dividing cells, bromodeoxyuridine (BrdU) and by immunostaining of post-mortem brains with antibodies against BrdU and neuronal markers. An endogenous marker of dividing cells, Ki-67 protein, has also been used instead of BrdU for this purpose. Thus, in healthy, young rodents, approximately 3,000-15,000 new cells per day are estimated to be born in the dentate gyrus of the hippocampus, about 60% of which express early neuron-specific proteins such as doublecortin and type III beta-tubulin. Significant numbers of new cells and new neurons have also been observed in healthy, young primates. In rodents as well as in primates, the location of neurogenic areas in the central nervous system (CNS) is limited to the dentate gyrus of the hippocampus and the subependymal layer of the striatum. In human patients of different ages who have been diagnosed with a tumor of the tongue, a single injection of BrdU has revealed significant number of new cells and new neurons being born in the dentate gyrus and the subependymal layer of the striatum. Thus, adult mammalian brains contain neural stem cells capable of differentiating into neurons, and this process of generating new neurons (neurogenesis) occurs in the mature, adult brain in significant quantities throughout rodents, primates, and human species.
Such significant quantities of new neurons suggest that the new neurons may be important for the normal physiology of the brain, especially the hippocampus. The hippocampus is the main area of neurogenesis in adult rodents and is central for key cognitive functions such as learning and memory where new information is added, edited, stored, and recalled constantly throughout life.
Since the hippocampus is the most potent neurogenic area of the brain, many studies have been undertaken to establish whether neurogenesis may be the cellular mechanism to structurally accommodate the ever-increasing volume of cognitive processing to be handled. Thus, it has been shown that at least some of the newly born neurons, marked by genetic markers, mature to be electrophysiologically active and integrate into the existing neuronal circuitry of the hippocampus. Ablation of the neurogenesis in rats leads to decreased cognitive capabilities in several behavior tests. Thus, the existing data demonstrate that neurogenesis significantly contributes to the normal hippocampal physiology.
For example, most antidepressants are thought to work by increasing the levels of monoamines available for post-synaptic receptors. Examples of classes of agents working apparently by the “monoaminergic hypothesis of expression” include the selective serotonin uptake inhibitors (SSRIs) like fluoxetine, the mixed noradrenaline/serotonin transporter blockers like tricyclic agent imipramine and noradrenaline uptake inhibitors like desipramine. The antidepressant-induced increase in intraneuronal biogenic amines occurs quite rapidly. However, the antidepressant-induced improvement in clinical behavior requires weeks of daily administration.
One hypothesis that may account for the slow-onset of the antidepressants' therapeutic activity is that they work by promoting hippocampal neurogenesis. It is expected that neurogenesis would require a number of weeks for stem cells to divide, differentiate, migrate and establish connections with post-synaptic neurons. The neurogenesis theory of depression then postulates that antidepressant effect is brought about by structural changes in the hippocampal circuitry contributed by newly generated neurons stimulated by antidepressants (Malberg et al., 2000; Czeh et al, 2001; Santarelli et al, 2003).
The neurogenic theory of depression, though not conclusive, has strong supportive data including the finding that neurogenesis is actually requisite for antidepressant behavioral improvement in the novelty suppressed feeding model (Santarelli et al., 2003). A therapeutic benefit from hippocampal neurogenesis is further supported by the finding of hippocampal atrophy in depression, where magnetic resonance imaging studies identified a reduction in the right and the left hippocampal volumes in individuals with major depression (Sheline et al., 1996; Bremner et al., 2000; Mervaola et al., 2000). Long-standing work also suggests a strong relationship between glucocorticoid dysregulation or glucocorticoid hypersecretion in stress and depression, such that the hippocampal volume loss might be considered a consequence of glucocorticoid-induced hippocampal neuronal loss (Sheline et al., 1996; Lucassen et al., 2001; Lee et al., 2002 (review)). Furthermore, in studies which involved the administration of a chronic stress to animals, both hippocampal volume changes and reduction in neurogenesis are observed, and these events are both reversed by chronic antidepressant administration (Czeh et al., 2001; Pham et al., 2003), further illustrating the strong association between depression, stress and neurogenesis. The association comes full circle, since agents or conditions that promote a reduction in neurogenesis also appear as causative agents/events in depression, specifically glucocorticoid (Sapolsky, 2000), and depletion of serotonin (Brezun and Daszuta, 1999). Kempermann and Kronenberg (2003), though acknowledging the validity of the hippocampal neurogenesis theory of depression, suggest that this hypothesis needs to be looked at in the context of a larger model of cellular plasticity, which elucidates how antidepressants induce nascent neurons of unknown phenotype to survive and produce viable circuits.
Neurogenesis can be characterized as three successive stages: proliferation of endogenous stem cells and precursors, differentiation into neurons and neuron maturation with formation of viable synaptic connections (plasticity). In consideration of each of these stages of neurogenesis, the hippocampal volume loss in depression could potentially be caused by 1) inhibition of the endogenous hippocampal stem cell proliferation in the dentate gyrus, 2) inhibition of differentiate and dendrite development and 3) loss of neurons (apoptosis) and their dendritic structure. Though apoptosis, also known as programmed cell death, is observed in depression, hippocampal apoptosis, as measured by DNA fragmentation, from depressed patients appears to play only a minor role in the volume loss (Lucassen et al., 2001).
In an animal model of acute stress or in normal animals receiving exogenous corticosterone, the stress causes a reduction in synaptic plasticity in the hippocampus (Xu et al., 1998). Chromic administration of the tricyclic antidepressant imipramine, partially reversed the loss in long-term potentiation (LTP) in socially stressed, depressive-like animals (Von Frijtag et al., 2001) suggesting imipramine affects the plasticity phase of neurogenesis.
In another animal model of depression characterized by loss of neurogenesis and hippocampal volume loss, stressed animals that chronically receive the antidepressant tianeptine, show similar numbers of dividing cells as control animals (no social stress) a measure of proliferation (Czeh et al., 2001).
In an experiment looking at association of antidepressants and neurogenesis in normal adult rats, the antidepressant, fluoxetine, required chronic administration to cause proliferation of cells in dentate gyrus (24 hrs post-treatment), but there was considerable loss of nascent cells, whether in the presence or absence of fluoxetine treatment, where fluoxetine provided no observed differentiation or survival benefit (Malberg et al., 2000).
Results on different neurogenic intervention points by known antidepressants suggest that novel neurogenic agents that intervene at different points in the neurogenesis pathway could result in potentially diverse therapeutic effects on depression. These points of intervention can be studied and the target elucidated for novel antidepressant candidates through in vitro assays. Since adult stem cell proliferation and neurogenesis is observed in adult vertebrates in hippocampal dentate gyrus (Gould el al., 2001; Eriksson et al., 1998), we can use multi-potential hippocampal stem cells screen agents in vitro for neurogenic activity.
Interestingly, chronic administration of either the antidepressant fluoxetine, an SSRI or the antidepressant rolipram, a phosphodiesterase IV inhibitor, promoted neurogenesis in normal animals (Malberg et al., 2000; Nakagawa et al., 2002). One might conclude from these results that any agent that promotes neurogenesis will generate a behavioral benefit in depression, unrelated to the agents' mechanism-of-action or possibly that there is a common pathway where both drug actions overlap. D'Sa and Duman suggest a scheme whereby the phosphorylation and activation of CREB and the subsequent expression of BDNF are central to the induction of neurogenesis that culminates in antidepressant behavior. CREB phosphorylation is increased in animals administered rolipram chronically (Nakagawa et al., 2002), and antidepressants that either increase Ca2+/CaM-kinases or cAMP could cause the phosphorylation of CREB in the nucleus (reviewed by D'sa and Duman 2002). D'Sa and Duman (2002) further suggest that the phosphorylated CREB then binds to the CRE binding site to promote the expression of BDNF and bcl-2, that appear critical to cell survival and plasticity. Proof of involvement of neurotrophic factor BDNF in depression comes from studies showing that antidepressants and electroconvulsive shock both caused an increase in BDNF levels (Nibuya et al, 1996) and that intrahippocampal injection of BDNF had antidepressant activity in two models of depression (Shirayama et al., 2002).
If neurogenesis is critical for antidepressant activity is it also sufficient for therapeutic activity and is the mechanism by which the neurogenesis occurs or timing of neurogenesis also critical to therapeutic activity? These questions can be answered by using novel agents developed through screening paradigms that identify agents that promote the proliferation and differentiation of endogenous hippocampal stem cells to neurons in vivo and determine if they will be effective antidepressants.
In abnormal conditions, such as when an injury to a brain area has occurred, neurogenesis becomes more wide-spread and perhaps functionally diverse. In rodent models of ischemic and hemorrhagic stroke, the newly born neurons of the subependyma (also referred to as subventricular zone) are seen migrating to and accumulating in the lesion area of the cortex. However, the newly born neurons have a short survival period.
Neuropathology associated with key cognitive regions of the Alzheimer's diseased brain has led to therapeutic strategies that address the neuronal loss, in the hopes of reducing the cognitive decline. One strategy enlists trophic agents, that regulate neuronal function and survival, as Alzheimer's Disease (AD) therapeutics (see Peterson and Gage, 1999). Problems with systemic administration, side effects and locating trophic-sensitive neurons generated few clinical successes with these therapies. One AD therapeutic, AIT-082, promotes memory enhancement in AD individuals potentially by stimulating endogenous trophic factors (Ritzman and Glasky, 1999; Rathbone et al., 1999). So the use of agents to promote increased survival and function of the remaining available neurons appears to have some therapeutic value.
As discussed, the hippocampus is one of the main brain regions where neurogenesis in adult brain has been documented across several vertebrate species, including monkeys and humans (e.g., Gould et al., 2001; Eriksson et al., 1998). In fact, adult hippocampal neurogenesis contributes functionally to cognitive capacity. Shors et al, (2001) reported that inhibition of neurogenesis in adult rat hippocampus, in the absence of the destruction of existing neurons, caused impaired memory function. Many studies observed that degenerative conditions induced, neurogenesis in mature mammalian brains, suggesting the existence of a natural repair pathway by means of neurogenesis. A focal ischemic model, reversible photothrombic ring stroke, caused increased neurogenesis in rat cortex by 3-6% (Gu et al., 2000). Seizures induced by electroconvulsive shock in adult rats increased neurogenesis in dentate gyrus of hippocampus (Scott et al, 2000; Madsen et al, 2000). Also, rats gamma-irradiated to kill mitotic cells in the CNS showed reduced numbers of nascent neurons and reduced LTP in slice cultures. These observations highlight the likelihood that a cellular mechanism for neurogenesis within adult human CNS, especially in hippocampus, does exist both as a normal physiological process and as a self-repairing pathway.
In adult neurogenesis a decline due to aging is observed (Kuhn et al., 1996), though proof that this age-dependent decline in neurogenesis causes cognitive impairment is still debated. Considerable evidence does exist, indicating that increased neurogenesis reduces age-associated cognitive decline. This is most dramatically observed with the transplantation of human stem cells into aged rats initiating improved water maze learning and retention (Qu et al., 2001). Other data suggests that induction of neurogenesis by diet restriction (Lee et al., 2000) exercise (van Prang et al., 1999) or growth factor addition (Lichtenwalner et al., 2001) improves learning and memory in adult or aged rats. A number of other inducers of neurogenesis have been identified, including anti-depressants (Malberg et al., 2000; Czeh et al., 2001), and nitric oxide donors (Zhang et al., 2001) suggesting the usefulness of neurogenic agents for other diseases presenting cognitive-deficits, such as stroke and depression. A small molecule that induces hippocampal neurogenesis that is blood brain barrier penetrable would allow for a potentially novel oral therapeutic for Alzheimer's disease.
Other potential AD therapeutics progressing in clinical trials, target neurodegeneration in the hopes of reducing the neuronal loss and cognitive decline. Apoptotic death involving caspase pathways and DNA fragmentation has been measured in in vitro and animal models of AD and in Alzheimer's diseased brain tissue. The extent of apoptosis leading to neuronal loss is of continual debate with most agreeing it has some effect, but that other neuronal death pathways definitely play a role (see Behl, 2000; Broe et al., 2001, Roth, 2001). Concern that measures of upstream caspase markers in neurons from AD tissue may not proceed to degeneration has been suggested (Raina et al, 2001). In order to screen for a neuroprotectant therapeutics it is critical, therefore, to measure more than one endpoint of neuronal death and determine at what point an agent may intervene in the death pathway(s). Behl (2000) suggested that AD pathology is most likely a mixture of apoptotic and necrotic pathways and that concentrating therapeutic discovery using only one pathway may provide inconclusive results. All hits in our neurogenesis models were tested through our secondary apoptosis/necrosis assay to screen for agents that function both as neurogenic and neuroprotective agents. These agents may improve or reverse the cognitive decline observed in MCI and AD.
Therefore, recent studies indicate that neurogenesis occurs in the adult human brains under normal as well as under degenerative conditions and that such adult-generated neurons do contribute functionally to the brain physiology such as learning and memory. These observations highlight the likelihood that a cellular mechanism for neurogenesis within adult human CNS, especially in hippocampus, does exist both as a normal physiological pathway and as a self-repairing pathway. What is not known is whether deficiencies in the volume or persistence of neurogenesis and/or the survival or maturation of the new neurons contribute to permanent damage.
Thus, a compound that can stimulate endogenous neurogenesis, either in a disease state or in a healthy state, may be an effective drug for a number of human nervous system conditions and diseases. Many neurological diseases, including Alzheimer's disease, mild cognitive impairment, dementia, age-related cognitive decline, stroke, traumatic brain injury, spinal cord injury, and the like, are neurodegenerative conditions. Neuropsychiatric diseases including depression, anxiety, schizophrenia and the like also show nerve cell dysfunction leading to cognitive, behavioral, and mood disorders. A neurogenic drug or agent that enhances the process of generating new neurons (neurogenesis) would be beneficial for countering and treating these diseases.
Candidate drugs generated from the screening have been tested in various animal models of human neurological and psychiatric disorders to determine the drugs' therapeutic potentials. An effective, predictive in vitro assay that can be used to select for clinical drug development neurogenic compounds that is particularly effective in promoting the neurogenesis in vivo has been described in U.S. patent application Ser. No. 10/728,652 filed Dec. 5, 2003, which is incorporated herein by reference in its entirety.